Neuroendocrine differentiation has been found in 50% to 100% of prostate neoplasms of all stages. Furthermore, 10% of prostatic tumors are extensively or fully differentiated. Thus, neuroendocrine-specific growth factors such as bombesin represent novel therapeutic targets and may also be important in the selection of patients for aggressive chemotherapy, which has been rather ineffective in prostate cancer. The use of more reliable prognostic tools or staging systems may overcome this difficulty and a nonhistological prostate cancer staging system has already shown that patients with visceral involvement but minimal bone disease have the highest chance of responding to a combination of doxorubicin, mitomycin-C, and 5-fluorouracil. Favorable response rates have also been demonstrated in trials of ketoconazole/Adriamycin or mitomycin-C combined with vinblastine/estramustine or vinblastine/Adriamycin.
Although hormonal treatment is an important palliative measure in the management of patients with advanced prostate carcinoma, the vast majority of cases end in relapse due to overgrowth of androgen-independent tumor cell populations . Recent studies have focused on neuroendocrine differentiation as a mechanism by which these tumors escape hormonal control [2-4]. This model of disease progression may be an important prognostic tool and have implications for the treatment of patients with stage D malignancy. This review will examine the evidence for the neuroendocrine progression of prostate cancer and the clinical utility of chemothera- peutic agents that are being investigated for the management of late stage disease.
Early data detected neuroendocrine cell types in only about 10% of prostate tumors analyzed [3 ]. However, the proportion of neuroendocrine differentiation was found to be much higher in follow-up studies and it is now thought that 50% to 100% of all prostate neoplasms (including benign prostatic tissue) contain at least some neuroendocrine cells [3,5]. It has been estimated that 10% of these tumors are extensively or fully differentiated. The latter are histologically similar to small cell or carcinoid tumors.
The availability of an anti-androgen receptor (AR) monoclonal antibody has allowed the identification of both AR-positive and AR-negative neuroendocrine subpopulations in both benign and malignant tumors . Combined with observations of neuroendocrine enrichment in cases of poor clinical response to androgen deprivation therapy, these findings support the notion that the clonal expansion of AR-negative neuroendocrine cells may play a critical role in the development of androgen-insensitive metastatic prostate cancer.
The clinical relevance of immunohistological detection of neuroendocrine tissue in prostate cancer patients has been the focus of many investigators.
Prognosis--Several studies have demonstrated that a large proportion of neuroendocrine differentiated prostatic neoplasms are positive for neuron-specific enolase and chromogranin A [4,6]. Preliminary data suggests that these markers may contribute to prostate cancer progression. Should this be confirmed, this may add to the routine histological grading in patients with progressive prostate cancer . Small-cell carcinomas are stimulated by bombesin/GRP and related peptides and contain receptors for them. Similarly, serum bombesin levels appear to be a useful prognostic tool in cases of neuroendocrine differentiated disease. Data obtained from patients with androgen- independent prostatic carcinoma have revealed that prostate-specific antigen (PSA) correlates with survival in patients with normal bombesin levels. Although indivi- duals with elevated serum bombesin also have elevated PSA, the latter is not a reliable prognostic indicator in these patients (unpublished observations). In addition, recently published case histories involving neuroendocrine differentiated small-cell or carcinoid prostatic tumors have reported normal levels of PSA as well as nearly normal acid phosphatase levels in patients experiencing rapid disease progression .
Overall, these data indicate that conventional prostate cancer prognostic tools such as PSA are inadequate to predict behavior of neoplasms as they become androgen-insensitive and neuroendocrine tumor populations expand (Figure 1). In these cases, markers such as neuron-specific enolase, chromogranin A, or bombesin are more suitable predictors of survival than prostatic volume.
Treatment--Aside from their use in prognosis, neuroendocrine cells and their growth factors represent good therapeutic targets. It has been observed that neuroendocrine cells secrete peptides such as serotonin, somatostatin, and bombesin, which stimulate tumor proliferation and lead to the biologically aggressive nature of neuroendocrine prostatic carcinoma [2,3]. Small cell carcinomas are also stimulated by bombesin/gastrin-releasing peptide-related peptides and contain receptors for them. Trials of anti-bombesin monoclonal antibodies as a treatment for small-cell lung cancer are already underway . At present, animal data indicate that bombesin/gastrin-releasing peptide inhibition via an antagonist (RC-3095) may have clinical utility for prostate cancer as well .
A more complete understanding of the contribution of neuroendocrine elements in malignant prostatic tissue may be important in the development and assessment of aggressive chemotherapeutic regimens. For instance, neuroendocrine markers may be helpful in the identification of those patients who have the best chance of benefiting from early combination chemotherapy that is unnecessary in the majority of prostate cancer patients [6,7].
Disease Staging--There has been growing concern about the variable efficacy of both single and combination chemotherapeutic strategies in patients with hormone refractory prostate cancer [10,11]. It has been suggested that at least part of the difficulty in assessing the effects of these treatments has been the diverse nature of prostate cancer patient populations. The use of more reliable prognostic indices and/or staging systems may overcome this problem and allow a more accurate stratification of drug efficacy by disease severity. In fact, a nonhistological prostate cancer staging system has already been in use for over a decade.
This staging system consists of four clinical categories (Table 1) . An early study classified 62 patients with hormonal refractory disease in this fashion who were treated with a combination of doxorubicin, mitomycin-C, and 5-fluorouracil (DMF). Results indicated that patients with less severe OI disease had a higher response rate (52%) and the VI group had a significantly higher response rate (88%, P less than 0.02) compared with the more serious stage II designations (33% for OII and VII). Responding patients (defined by improvements in bone scans, blastic healing, and lab parameters as well as a more than 50% reduction in the sum of tumor diameters) had a significantly prolonged survival rate compared with nonresponders (47.5 vs 23.8 weeks, P = 0.001).
In addition to predicting survival benefit after chemotherapy, the staging system also predicted response duration . Patients with osseous disease had longer median durations of response (11 and 9.5 months for stages OI and OII, respectively) than those classified as VI (6.5 months) or VII (5 months). These data demonstrate that patients with visceral involvement but minimal bone disease have the highest chance of responding to DMF chemotherapy, while fewer patients with axial and distal skeletal involvement or visceral metastasis outside the lung experienced clinical benefits.
Bone Progression Assay--An in vitro assay for assessing the effect of bone on the progression of androgen-independent prostate tumor growth has recently been developed at the University of Texas. The acellular supernatant derived from patient bone marrow aspirate is added to confluent human prostate cancer cell cultures (LNCaP). In order to evaluate the capacity of bone marrow aspirate to induce in vitro PSA production, serum-free medium from these cultures is analyzed by Northern blot. Since androgens are thought to be the only method of regulating PSA release, it is assumed that induction of PSA by the supernatant should not occur in patients with castrate levels of testosterone. Any in vitro induction would correspond to a bone-specific disease process.
Using this methodology, Logothetis et al found that bone marrow aspirate from patients with localized, androgen-dependent disease, or small-cell carcinoma did not induce LNCaP PSA, presumably because disease progression is not dependent upon bone in these cases (unpublished observation, Figure 2). In contrast, in vitro PSA release occurred most frequently and intensely with bone marrow aspirate obtained from patients with osseous-dependent tumor progression as well as androgen-independent disease. Since it focuses on patients with metastatic lesions, the assay is capable of assessing the antitumor activity of therapies often administered during advanced disease (ie, external beam radiotherapy, strontium-89, aggressive chemotherapy).
Dosing schedule--Combination chemotherapeutic regimens are often administered weekly or monthly. Differ- ences in efficacy between mitomycin-C combined with weekly vinblastine/estramustine or monthly vinblastine/Adriamycin (Table 2) may be more depend- ent upon the dosing schedules than inherent differences in drug efficacy. It is possible that increases in clinical benefits derived from weekly schedules are related to effects on the deregulation of programmed tumor cell death (apoptosis) that is known to occur during prostate neoplastic development .
Vinblastine/Estramustine--Several groups have evaluated the clinical potential of combination chemotherapy with vinblastine (Velban) and estramustine (Emcyt) (VE), two microtubule inhibitors with independent mechanisms of action that appear to provide additive therapeutic effects [13-15]. Preliminary data obtained from 29 patients with hormone refractory prostate cancer indicated that oral estramustine (140 mg) taken three times daily and weekly intravenous bolus vinblastine (6 mg/m²) provided a clinical response rate of 24% to 50% (as measured by more than 50% reductions in PAP or PSA, respectively) . The overall objective response rate was 30%.
Later studies employing different VE dosing (ie, 4 mg/m² weekly vinblastine, 10 mg/kg or 600 mg/m² estramustine) found 54% to 61% rates of PSA reduction [13,15]. The median duration of response was seven months. In addition, the rates of subjective pain palliation in these studies were 43% to 66%. The estimated overall response rate provided by VE was 58% . Using the same dose and evaluation criteria, estramustine alone produced a clinical response rate of 29% while vinblastine monotherapy resulted in a 21% objective response rate in other studies. These data suggest at least an additive clinical effect of VE treatment. Logothetis et al are involved in ongoing prospective randomized trials to determine the relative contributions of each agent to clinical response during combination therapy.
Vinblastine/Estramustine/Mitomycin-C--Based on their previous clinical experience with VE, Logothetis et al have recently reported that the addition of mitomycin-C to the VE regimen substantially improves efficacy . The rate of PSA reduction among 31 patients with initially elevated PSA was 66% and the duration of response was 2 months. In addition, 7% of 28 patients with osseous disease experienced improved bone scans.
Suramin--Suramin (an agent that has been shown to inactivate fibroblast, platelet-derived, and insulin-like growth factors as well as interleukin-2) has also shown clinical potential in patients with hormone-refractory prostatic carcinoma . A recent trial involving 38 patients revealed that suramin treatment resulted in a reduction (by 75% of pretreatment value) or a normalization of pretreatment PSA in a total of 35% of patients with soft tissue disease. This PSA response was more frequent (64%) among patients with metastasis limited to bone. These improvements significantly correlated with increased survival probability among patients with either soft tissue (P = 0.0184) or bone disease (P = 0.016). Pain palliation was also experienced by 71% of patients who were previously receiving opiates and who could reduce or stop analgesia due to suramin.
The number of published clinical trials involving monotherapy or combination treatment with a wide variety of chemotherapeutic agents is too large to be included in this review . As new therapeutically exploitable steps in prostate cancer disease progression become elucidated, novel agents will likely be added to the armamentarium. For now, new methodologies are being developed to more fully characterize androgen independent neoplastic growth, neuroendocrine differentiation and enrichment, and the contribution of host properties to late stage disease progression. Advancements in these areas may allow us to maximize the efficacy of current chemotherapeutic protocols.
1. Denis L: Prostate cancer. Primary hormonal treatment. Cancer 71:1050-1058, 1993.
2. di Sant'Agnese PA, de Mesy Jensen KL: Neuroendocrine differentiation in prostatic carcinoma. Hum Pathol 18:849-856, 1987.
3. di Sant'Agnese PA: Neuroendocrine differentiation and prostatic carcinoma. Arch Pathol Lab Med 112:1097-1099, 1988.
4. Turbat-Herrera EA, Herrera GA, Gore I, et al: Neuroendocrine differentiation in prostatic ca. Arch Pathol Lab Med 112:1100-1105, 1988.
5. Nakada SY, di Sant'Agnese PA, Moynes RA, et al: The androgen receptor status of neuroendocrine cells in human benign and malignant prostatic tissue. Cancer Res 53:1967, 1993.
6. Cohen RJ, Glezerson G, Haffejee Z: Neuro-endocrine cells-a new prognostic parameter in prostate cancer. Br J Urol 68:258-262, 1991.
7. Smith DC, Tucker JA, Trump DL: Hypercalcemia and neuro-endocrine carcinoma of the prostate: A report of three cases and a review of the literature. J Clin Oncol 10:499-505, 1992.
8. Mulshine JL, Avis I, Treston AM, et al. Clinical use of a monoclonal antibody to bombesin-like peptide in patients with lung cancer. Annal NY Acad Sci 360-372, 1988.
9. Pinski J, Halmos G, Schally AV: Somatostatin analog RC-160 and bombesin/gastrin-releasing peptide antagonist RC-3095 inhibit growth of androgen-independent DU-145 human prostate cancer line in nude mice. Cancer Lett 71:189-196, 1993.
10. Petrylak DP, Scher HI, Li A, et al: Prognostic factors for survival of patients with bidimensionally measurable metastatic hormone- refractory prostatic cancer treated with single-agent chemotherapy. Cancer 70:2870-2878, 1992.
11. Logothetis CJ, Samuels ML, von Eschenbach AC, et al: Doxorubicin, mitomycin-C, and 5-fluorouracil (DMF) in the treatment of metastatic hormonal refractory adenocarcinoma of the prostate, with a note on the staging of metastatic prostate cancer. J Clin Oncol 1:368-379, 1983.
12. Yagoda A, Olsson C: Neoplasms of the kidney, bladder, and prostate, in Calabresi P, Schein PS, eds. Medical Oncology, 2nd ed. New York: McGraw-Hill, 1993.
13. Seidman AD, Scher HI, Petrylak D, et al: Estramustine and vinblastine: Use of prostate specific antigen as a clinical trial end point for hormone refractory prostatic cancer. J Urol 147:931-934, 1992.
14. Amato RJ, Logothetis CJ, Dexeus FH, et al: Preliminary results of a phase II trial of estramustine (Emcyt) and vinblastine (VLB) for patients with progressive hormone refractory prostate carcinoma (abstract 1111). Proc Am Assoc Cancer Res 1991;32:186, 1991.
15. Hudes GR, Greenberg R, Krigel RL, et al: Phase II study of estramustine and vinblastine, two microtubule inhibitors, in hormone-refractory prostate cancer. J Clin Oncol 10:1754, 1992.
16. Amato R, Logothetis C, Sella A, et al: Preliminary results of a phase II trial of estramustine (Emcyt), vinblastine (VLB) and mitomycin-C (MMC) for patients with progressive androgen independent prostate carcinoma (AIPCa) (abstract 1213). Proc Am Assoc Cancer Res 34:203, 1993.
17. Myers C, Cooper M, Stein C, et al. Suramin: A novel growth factor antagonist with activity in hormone-refractory metastatic prostate cancer. J Clin Oncol 10:881-889, 1992.